@dolphin_jay: Assuming the cross-function dependencies aren't totally over the top, it should also be possible to just remodel the existing code into a tight pack without the overhead. It seems to me that the bulk of the performance hit is brought about by what's essentially clutter in the context.

However, there's little point in the community preparing ready fixes for issues like this, if the software developers themselves can't see the problem. Give a man a fish, and all that -- would be sound see more if considerations like this were actually an integral part of their development approach.